Chronic illnesses in childhood affect many aspects of development. While insulin dependent diabetes mellitus (IDDM) has not been found to be associated with general intelligence, it is now evident that children with IDDM have difficulties in a number of specific cognitive tasks. Further, early age of onset of the disease contributes to the problems. Early onset diabetes (EOD) is clearly associated with poorer spatial skill. In each of the three studies on the topic, EOD children (i.e. by age 4 or earlier) have had lower spatial scores than children who developed IDDM later (LOD) or nondiabetic comparison children. These results are difficult to interpret, however, each of the studies relied exclusively upon psychometric measures. These measures are useful in describing an individual's position relative to a normative sample, and whether that individual's performance is impaired. However, they cannot reveal processes or strategies that are responsible for group differences. These differences in psychometric performance could reflect differences in use of these strategies or one group (e.g. the EOD group) may simply execute some or all of the component processes more slowly. This revised proposal is designed to determine whether, compared to nondiabetic children and children who develop diabetes later in life, children who develop diabetes early in life (a) rely upon different strategies for spatial tasks, or (b) use similar strategies but execute the component processes or more slowly. To evaluate these explanations, children and adolescents from EOD, LOD, and nondiabetic comparison groups will be tested on a series of spatial tasks in order to assess specifically the strategies used and the speed with which component processes are executed. Subjects will also be tested on non-spatial tasks. Their performance on these tasks will allow us to determine if findings concerning strategies and speeds of cognitive processing are unique to the spatial domain or hold more generally. Specifically, the pattern of response times of EOD, LOD, and comparison individuals will be examined to reveal the strategies used to perform the tasks. If the evidence indicates that the groups use the same strategy for solving the task, then the time to execute the processes that constitute the task will be compared. The sample will include 72 children, 24 children in each of three groups, ranging from 9-16 years of age. All will be generally healthy and attending regular school classroom. The EOD group will have been diagnosed with diabetes at age 4 years or younger. The LOD group will be those children diagnosed with diabetes after 4 years of age. Children in the comparison groups will be matched by age (within 6 months), sex, and grade in school with the EOD and LOD children. All children will be tested on the following tasks: Mental rotation, Form board, Geometric analogies, Visual search, two memory tasks, and measures of academic achievement and intelligence. Performance on the spatial and memory tasks will also be compared to academic achievement and estimates of general intelligence to determine if findings from earlier studies will be replicated.